Fuzzy adaptive control method for lower limb exoskeleton based on nonlinear disturbance observer

Abstract

The present invention discloses a fuzzy adaptive control method for lower-limb exoskeletons based on a nonlinear disturbance observer, applicable to the field of exoskeleton robots. Because accurate numerical values ​​for exoskeleton robot dynamic parameters are difficult to obtain, and external disturbances present during exoskeleton control are difficult to estimate, to address these issues and enable the exoskeleton to better track a set trajectory, a fuzzy system is used to estimate the unknown dynamic parameters, and a nonlinear disturbance observer is used to estimate lumped disturbances to compensate for the effects of uncertain disturbances. The controller employed in the present invention is used to drive the motors of the exoskeleton device, effectively improving the exoskeleton's responsiveness and tracking accuracy.

Description

Technical Field

[0001] The present invention belongs to the field of exoskeleton robots, and in particular relates to a lower limb exoskeleton control technology. Background Art

[0002] Exoskeleton robots, a typical wearable robot, form a human-machine coupling system when combined with the human body. They combine the robot's high mechanical strength and high load capacity with the human's environmental perception and task analysis capabilities, and have broad application value in medical rehabilitation, industrial production, and elderly and disability assistance. Lower-limb exoskeleton robots, as walking-assistance devices, couple the exoskeleton's mechanical structure with the human leg. Through human control and external power supply, they enable operators with limited or no mobility to walk autonomously. Different gaits and walking speeds can be designed to accommodate different patient conditions.

[0003] The lower limb exoskeleton mainly consists of the following parts:

[0004] (1) Mechanical structure: Lower limb exoskeletons usually adopt a hip + knee + ankle or hip + knee structure, depending on their functional requirements. For example, rehabilitation exoskeleton robots, which are mostly used for patients and require reduced joint movement, often adopt the latter structure. The materials used in the mechanical structure of the exoskeleton should have the characteristics of light weight, high strength, and fatigue resistance, such as carbon fiber, aluminum alloy, titanium alloy, and nanomaterials.

[0005] (2) Power system: This system mainly provides the power source for the exoskeleton's assistance, which can be provided by hydraulic pressure, air pressure, or motors. If a motor drive is used, it drives the device to complete the corresponding human-machine coupling task according to real-time control instructions.

[0006] (3) Sensor system: The exoskeleton's sensor system is primarily used to measure and sense the real-time operating status of the exoskeleton prototype, obtain various signals during human-machine interaction, determine the human gait or movement intention, and formulate exoskeleton control strategies and algorithms. Commonly used sensors include three-dimensional force sensors, absolute encoders, and torque sensors.

[0007] (4) Control system: The proposed control algorithm and related methods are usually implemented using software such as Matlab / Simulink and then downloaded to the corresponding hardware controller.

[0008] The existing technology currently has external disturbance influences, which makes the controller's response ability and tracking accuracy low, which is not conducive to assisting the wearer to complete specific tasks. Summary of the Invention

[0009] To solve the above technical problems, the present invention proposes a fuzzy adaptive control method for lower limb exoskeleton based on a nonlinear disturbance observer. The fuzzy system is used to estimate the uncertain dynamic parameters, and the lumped disturbance is compensated by the designed nonlinear disturbance observer. This can effectively reduce the adverse effects of external disturbances, enabling the exoskeleton robot to better track the set trajectory and complete the human-machine coupling task.

[0010] The technical solution adopted by the present invention is: a fuzzy adaptive control method for lower limb exoskeleton based on nonlinear disturbance observer, the control system based on which includes: a fuzzy adaptive controller based on disturbance observer, an exoskeleton model, a force sensor, and an admittance controller;

[0011] Force sensors are installed on the exoskeleton model to obtain the interaction force between the human body and the lower limb exoskeleton model. The interaction force obtained by the force sensor serves as the input of the admittance controller, and the output of the admittance controller is the trajectory correction value. The trajectory correction value is used to correct the reference trajectory to obtain the desired trajectory. The desired trajectory and the dual-joint angle of the exoskeleton serve as the input of the fuzzy adaptive controller based on the disturbance observer. The exoskeleton model moves under the control of the fuzzy adaptive controller based on the disturbance observer.

[0012] The process of obtaining a fuzzy adaptive controller based on a disturbance observer includes the following steps:

[0013] S1. According to Lagrangian dynamics, the 2-DOF lower limb exoskeleton dynamics model considering human-machine coupling can be expressed as:

[0014]

[0015] Among them, q represents the double joint angle of the exoskeleton, represents the joint angles of the exoskeleton, represents the angular acceleration of the exoskeleton, τ represents the driving torque of the motor, τ ext Represents the human-machine coupling torque, M(q), and G(q) represent the inertia matrix, Coriolis matrix and gravity term of the system respectively, f dis (t) represents the lumped term composed of unknown interference outside the system;

[0016] S2, compensate the lumped term by using a fuzzy adaptive controller based on disturbance observer, and solve M(q), and G(q); the fuzzy adaptive controller based on disturbance observer is expressed as:

[0017]

[0018]

[0019]

[0020]

[0021] Where, z 2,i is an element in z2, S(Z) is the fuzzy membership function, Γ i is a positive real number, represents the observation of interference, For a fuzzy system, Approximation Θ *T S(Z), Θ *T S(Z) is defined as:

[0022]

[0023] Among them, ε(Z) is the fuzzy approximation error, which is the lumped term f dis components of is the input variable of the fuzzy system

[0024] Satisfy the following formula:

[0025]

[0026] Beneficial effects of the present invention: The present invention provides a fuzzy adaptive control method for lower limb exoskeleton based on a nonlinear disturbance observer; the designed nonlinear disturbance observer can effectively estimate external disturbances and reduce the impact thereof; the fuzzy adaptive controller used approximates uncertain dynamic parameters, and can achieve stable control of the lower limb exoskeleton system to ensure that it completes the designated task. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is the control block diagram of the exoskeleton designed by the present invention;

[0028] Figure 2 This is the basic structure diagram of the fuzzy control used in the present invention;

[0029] Figure 3 is the observation effect of the interference observer of the present invention;

[0030] Among them, (a) is the hip joint, (b) is the knee joint;

[0031] Figure 4 This is a comparison chart of the tracking effect and tracking error of the controller of the present invention;

[0032] Among them, (a) is the hip joint tracking effect, (b) is the knee joint tracking effect, (c) is the hip joint tracking error, and (d) is the knee joint tracking error. DETAILED DESCRIPTION

[0033] To facilitate those skilled in the art to understand the technical content of the present invention, the present invention is further explained below with reference to the accompanying drawings.

[0034] like Figure 1 The control block diagram shown in the figure is mainly composed of a position controller, an exoskeleton model, a human body model, a force sensor, and an admittance controller. The admittance controller can achieve the effect of human-machine following. Its input force information is the interaction force between the human and the machine, which can be obtained through the three-dimensional force sensor installed on the exoskeleton. The trajectory correction value △q output by the admittance model depends on the admittance parameters and the interaction force between the human and the machine, reflecting the human body's movement intention. △q is used to adjust the reference trajectory q r Make corrections to obtain the true desired trajectory q of the position controller d .Depend on Figure 1 It can be seen that the admittance controller needs to be used in combination with the position controller. At the same time, existing experience also shows that the performance of the position controller largely determines the performance and stability of the admittance controller.

[0035] In this embodiment, the position controller selects the fuzzy adaptive controller based on the nonlinear disturbance observer designed in the present invention. The nonlinear disturbance observer can make a prediction of the external disturbance within a certain fault tolerance range and compensate the lumped term f when designing the control rate. dis (t), improving the control accuracy of the exoskeleton. The fuzzy adaptive controller is used to reduce the impact of the exoskeleton's uncertain dynamic parameters and ensure that the exoskeleton tracks the ideal trajectory.

[0036] In this embodiment, the specific design steps of the controller are:

[0037] According to Lagrangian dynamics, the 2-DOF lower limb exoskeleton dynamic model considering human-machine coupling can be expressed as:

[0038]

[0039] Where, represents the dual joint angle, angular velocity and angular acceleration of the exoskeleton, R represents a set of real numbers, τ∈R 2 Represents the driving torque of the motor, τ ext ∈R 2 Represents the human-machine coupling torque, M(q), and G(q) represent the inertia matrix, Coriolis matrix and gravity term of the system respectively, f dis (t) represents the lumped term composed of unknown interference outside the system. and G(q) contain unknown kinetic parameters.

[0040] In this embodiment, the state variables of the exoskeleton system are set to x1 = [q1, q2] T , According to the dynamic model, the state space expression of the system is defined as:

[0041]

[0042]

[0043] The T in the superscript represents matrix transpose, q1 represents the hip joint angle, q2 represents the knee joint angle, is the hip joint angular velocity, represents the knee joint angular velocity. In this embodiment, adding a dot in the superscript indicates the first derivative, and adding two dots indicates the second derivative, for example It means to find the first derivative of q, Indicates the second derivative of q; M is the abbreviation of M(q), M -1 Indicates the inversion of M; G is the abbreviation of G(q); C is abbreviation of ;

[0044] The state error of the exoskeleton is defined as: z1 = x1 - x d , z2=x2-α.

[0045] Among them, x d It is represented as the set trajectory, and α is the virtual control amount. According to the two-degree-of-freedom exoskeleton robot, we can get

[0046]

[0047] Design of Lyapunov function The time derivative of V1 is:

[0048]

[0049] You can get:

[0050]

[0051] in, Design of Lyapunov function The time derivative of V2 is:

[0052]

[0053] If all the dynamic model parameters are known, the backstepping controller is designed as:

[0054]

[0055] You can get:

[0056]

[0057] K1 and K2 represent gain matrices;

[0058] Therefore, the state errors z1 and z2 of the exoskeleton system tend to 0 when t→∞, that is, the system is globally asymptotically stable. However, it is difficult to obtain the exact lumped term f dis As well as the robot dynamic model parameters G, C, M, a fuzzy system can be used to solve the uncertain dynamic model parameters. The present invention designs a nonlinear disturbance observer to compensate for the lumped term. The designed fuzzy adaptive controller is:

[0059]

[0060]

[0061]

[0062]

[0063] Where, Contains approximator parameters, := indicates the definition, blockdiag indicates the matrix block, n indicates the number of elements contained, yes The components of z 2,i ∈R is an element in z2, is the fuzzy membership function, S k It's S k (Z) abbreviation, is the input variable of the fuzzy system, q ik represents the element in row i and column k of the fuzzy input Z, c ik and σ ik Represented respectively The width of the center and the border; Γ i is a positive real number. In the present invention, let Fuzzy systems Approximation Θ *T S(Z), Θ *T S(Z) is defined as:

[0064]

[0065] Where ε(Z)∈R 2 is the fuzzy approximation error, which is also the lumped term f dis Components of.

[0066] In this embodiment, the disturbance observer compares the actual output of the system with the estimated output and uses the error to correct the estimation of the system in real time.

[0067]

[0068] in, Represents the observation of the interference, and f represents the unknown external interference. Usually in the actual process, it is generally impossible to know the prior information of the differential of the interference term, but the change of the interference is often slow, so the differential term can be assumed to be 0, that is,

[0069]

[0070] Assume that the observation error is

[0071]

[0072] The time derivative of the observation error is:

[0073]

[0074] In practice, it is difficult to obtain an accurate acceleration signal by differentiating the robot's velocity signal, so a more practical nonlinear disturbance observer needs to be designed. Design auxiliary vectors:

[0075]

[0076] Among them, β∈R 2 , is a nonlinear vector, and the derivative of the auxiliary vector β can be obtained:

[0077]

[0078] In formulas (12), (15), and (18), It is given by:

[0079]

[0080] We can get:

[0081]

[0082] Therefore, the nonlinear disturbance observer can be designed as follows:

[0083]

[0084] because

[0085]

[0086] Therefore, the function can be designed The observation error index is made close to 0. Since the nonlinear observer does not require the accurate dynamic parameters of the model, the present invention uses nominal parameters to design the disturbance observer.

[0087] In this embodiment, for the designed interference observer, when the function in the observer is At this time, the observer is globally asymptotically stable; c represents the maximum velocity of the knee joint.

[0088] like Figure 3 As shown in FIG, the tracking effect of the nonlinear disturbance observer of the present invention on the external lumped disturbance is given, and it can be seen that the general trend can be tracked. Figure 4 As shown in FIG, the control effect and tracking error with and without the observer are compared. It can be seen that the fuzzy adaptive controller based on nonlinear disturbance observer (NDO) of the present invention has better control effect than that without the observer (NONDO).

[0089] Figure 3 Where f(t) represents the external aggregate interference, Figure 4 Here, Angle response (deg) indicates angle response (degrees); Time indicates time.

[0090] Those skilled in the art will appreciate that the embodiments described herein are intended to aid the reader in understanding the principles of the present invention, and it should be understood that the scope of the present invention is not limited to such specific descriptions and embodiments. Various modifications and variations are readily apparent to those skilled in the art. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention are intended to be included within the scope of the claims.

Claims

1. A fuzzy adaptive control method for lower limb exoskeleton based on nonlinear disturbance observer, characterized in that: The control system it is based on includes: fuzzy adaptive controller based on disturbance observer, exoskeleton model, force sensor, and admittance controller; Force sensors are installed on the exoskeleton model to obtain the interaction force between the human body and the lower limb exoskeleton model. The interaction force obtained by the force sensor serves as the input of the admittance controller, and the output of the admittance controller is the trajectory correction value. The trajectory correction value is used to correct the reference trajectory to obtain the desired trajectory. The desired trajectory and the dual-joint angle of the exoskeleton serve as the input of the fuzzy adaptive controller based on the disturbance observer. The exoskeleton model moves under the control of the fuzzy adaptive controller based on the disturbance observer. The process of obtaining a fuzzy adaptive controller based on a disturbance observer includes the following steps: S1. According to Lagrangian dynamics, the 2-DOF lower limb exoskeleton dynamics model considering human-machine coupling can be expressed as: Among them, q represents the double joint angle of the exoskeleton, represents the dual-joint angular velocity of the exoskeleton, represents the dual-joint angular acceleration of the exoskeleton, τ represents the driving torque of the motor, τ ext Represents the human-machine coupling torque, M(q), and G(q) represent the inertia matrix, Coriolis matrix and gravity term of the system respectively, f dis (t) represents the lumped term composed of unknown interference outside the system; S2, compensate the lumped term by using a fuzzy adaptive controller based on disturbance observer, and solve M(q), and G(q); the fuzzy adaptive controller based on disturbance observer is expressed as: Where z1 and z2 are the state errors of the exoskeleton system, K2 represents the gain matrix, To include the approximator parameters, := means it is defined as, blockdiag means the matrix block, n means the number of elements included, yes The components of z 2i is an element in z2, S(Z) is the fuzzy membership function, Γ i is a positive real number, is the design function, β is the auxiliary vector, is the first derivative of β, is the function in the observer, represents the observation of interference, It is a fuzzy system.

2. The fuzzy adaptive control method for lower limb exoskeleton based on nonlinear disturbance observer according to claim 1 is characterized in that: Approximation Θ *T S(Z), Θ *T S(Z) is defined as: Among them, α is the virtual control quantity, is the first derivative of α, ε(Z) is the fuzzy approximation error, and is the lumped term f dis components of is the input variable of the fuzzy system.

Images